Brent Fulgham <[EMAIL PROTECTED]> writes: > One other thing that is probably not a problem, but > caught my eye, is the way the conditional operator > is used here: > > + ? (sblock->s_prealloc_blocks ?: EXT2_DEFAULT_PREALLOC_BLOCKS) > ^^^ > Will this construct (?:) result in the conditional passing > a NULL to the ext2_new_block method? On IBM's compiler > for AIX this won't even compile.
This construct is a GNU C extension. a ?: b is equivalent to a ? a : b, except that a is evaluated only once. I.e. ?: behaves the way || would have behaved if C did the Right Thing ;) Compare with lisp's (or a b) . /Niels
